Serial electrocardiographic findings in a prospective epidemiological study.

THE PREVALENCE of a number of elec-trocardiographic findings among the adult population of a total natural community was reported in an earlier article.' This community , Tecumseh, Michigan, is the site of a long-term study of a great variety of factors which may influence the health of the population. The first series of complete examinations of the participants was carried out in 1959 and 1960. A second series of examinations of the Tecumseh population was completed in June 1965. It is now possible to compare the prevalence of electrocardiographic findings among most of the adult members of a single community at two periods about 4 years apart and to analyze the changes which have occurred among that segment of the population who had electrocardiograms during both examinations. This report examines the problem of re-producibility of electrocardiographic classifications when a uniform coding system is used by different cardiologists and gives information about the stability of certain electrocar-diographic abnormalities over a period of several years. It also allows an estimate of the value of nonspecific electrocardiographic changes in the prediction of more definite abnormalities. Electrocardiographic findings rather than specific diagnoses are discussed. Diagnoses are based on information from the medical history , physical examination, and sometimes postmortem reports as well as the electrocar-diogram. The incidence of new cardiovas-cular disease and the relationship of the various conditions to antecedent risk factors will be discussed in detail in subsequent reports. Methods Electrocardiograms were recorded from 5,129 examined persons 16 years of age and over in 1959 and 1960. During the second period of examinations from 1962 to 1965 tracings were obtained from 5,451 persons. Of the persons from whom electrocardiograms were recorded and classified in 1959 and 1960, 3,745 or 73% had a second tracing between 1962 and 1965. The percentage of participants who had an electrocardiogram on both examinations was considerably higher, but few of the tracings from the 16 to 21-year-old group were classified according to the adult criteria on their first examination because they were less than 16 years of age. The interval between examinations varied from 20 to 72 months with a median interval of 47 months. Eighty per cent of the participants were reexamined between 40 and 56 months after their first examination. The electrocardiograms were recorded on direct writing single-channel instruments at a paper speed of 50 mm per second. Food and tobacco were not prohibited before the …